py-concodese 3.1.0

Given a bug report, suggests code files that may need to be fixed

README

PyConCodeSe is a bug localisation tool: given a bug report it suggests 10 code files that may contain the bug.

For PyConCodeSe to work it needs a Tree-sitter grammar to parse your code (we currently support Java, C#, PHP and Rust) and the spaCy library to parse the bug report and any natural language text in the code.

After installing the software as per the following instructions, you will need to edit a configuration file for each code repo. The configuration tells PyConCodeSe where your code and your bug reports are.

Installation

You need at least Python 3.11 to run PyConCodeSe and we strongly recommend you install it in a virtual environment. On Unix (incl. macOS) you would type in a terminal something like:

python3.11 -m venv venv
source venv/bin/activate

to create and activate a virtual environment in subfolder venv of your current folder. Next, type

pip install -i py-concodese

to install everything in the current environment and create the main executable script pyconcodese. Finally, set up spaCy with

python -m spacy download en_core_web_sm

Now you can run pyconcodese from any folder, as long as you have the virtual environment activated.

Configuration

When running PyConCodeSe, if there's no file config.toml in the current folder, you're prompted to create one with example values.

pyconcodese 

The PyConcodese config file 'config.toml' was not found in this path. 
 Do you want to copy an example file here? [y/N]y
Example config.toml file copied in the current directory.  
 Please make sure you change the fields accordingly

Open the configuration file and edit the paths of the various folders:

[py-concodese]
# the next 3 can be the root directory of the application if you wish (e.g.)
# the directory that you would like to store the sqlite database(s) in
sqlite_path = "/tmp/Pyconcodese"
# the directory that you would like to store the vsm data folders in
vsm_path = "/tmp/Pyconcodese"
# where to store output files generated by the application
output_path = "/tmp/Pyconcodese"
derby_path = "/tmp/Pyconcodese/derby"


[dataset]
bug_repository_file = "/tmp/Pyconcodese/bug_repository_file.xml"
src_path = "/tmp/Pyconcodese"

The config.toml file will be created with the current path prefilled as a starting point for all these values.

Dataset format

The dataset (bug reports) that you want to use should be in a particular format. Necessary fields that should be present in the dataset are: issue_id, issue_summary, issue_description, issue_status and files_changed.

The following examples show the correct format in which the dataset should be aligned:

Json version

{
    "closed_issues": {
        "1": {
            "issue_id": "#3085",
            "issue_summary": "Missing PR_SET_PTRACER_ANY",
            "issue_description": "JonathanWoollett-Light…..search=PR_SET_PTRACER_ANY).",
            "issue_status": "Closed",
            "files_changed": [
                [
                    "1",
                    "libc-test/semver/fuchsia.txt"
                ],
                [
                    "1",
                    "libc-test/semver/linux.txt"
                ],
                [ 
                    "1", 
                    "src/fuchsia/mod.rs" 
                ], 
                [ 
                    "1", 
                    "src/unix/linux_like/emscripten/mod.rs" 
                ], 
                [ 
                    "1", 
                    "src/unix/linux_like/linux/mod.rs" 
                ] 
            ] 
        }, 
        "437": { 
            "issue_id": "#100", 
            "issue_summary": "Outdated MIPS toolchain", 
            "issue_description": "Contributor…..further.", 
            "issue_status": "Closed", 
            "issue_reporting_time": "", 
            "fixed_by": "#114", 
            "files_changed": [ 
                [ 
                    "1", 
                    "src/unix/notbsd/linux/mips.rs" 
                ], 
                [ 
                    "1", 
                    "src/unix/notbsd/linux/mod.rs" 
                ], 
                [ 
                    "1", 
                    "src/unix/notbsd/linux/musl.rs" 
                ], 
                [ 
                    "1", 
                    "src/unix/notbsd/linux/other/mod.rs" 
                ] 
            ] 
        } 
    } 
} 

XML version

<?xml version="1.0" encoding="ISO-8859-1"?>

<bugrepository name="SWT">
  <bug id="88829" opendate="2005-03-22 20:41:00" fixdate="2005-04-06 17:05:00">
    <buginformation>
      <summary>Table.setColumnOrder() may not fire enough Move events</summary>
      <description>- start with five columns, all different widths - do Table.setColumnOrder(new int[] {4,1,2,3,0}); - SWT.Move events are fired for columns 0 and 4 because they swapped positions -&amp;gt; but Move should have been fired for all of the columns since the width of the first displayed column changed, and therefore all of the other columns are auto-shifted accordingly</description>
    </buginformation>
    <fixedFiles>
      <file>org.eclipse.swt.widgets.Table.java</file>
    </fixedFiles>
  </bug>
  <bug id="90018" opendate="2005-04-01 14:40:00" fixdate="2005-04-05 08:14:00">
    <buginformation>
      <summary>Native tooltips left around on CTabFolder</summary>
      <description>Hover over the PartStack CTabFolder inside eclipse until some native tooltip is displayed. For example, the maximize button. When the tooltip appears, change perspectives using the keybinding. the CTabFolder gets hidden, but its tooltip is permanently displayed and never goes away. Even if that CTabFolder is disposed (I'm assuming) when the perspective is closed.</description>
    </buginformation>
    <fixedFiles>
      <file>org.eclipse.swt.custom.CTabFolder.java</file>
    </fixedFiles>
  </bug>
</bugrepository>